Printable View
you SHOULD be looking out for Conroe :) , similar to Yonah but designed from the ground up for desktop use with 64bit enabled :D (please correct me if i am wrong)Quote:
Originally Posted by Northwood
also, regarding intel regaining enthusiast populatirty all in one fell swoop :rolleyes: , i VERY much doubt it, intel have become Sooooo good at snatching defeat from the jaws of success to start changing overnight :stick:
I saw this on tech report the other day, I have to say it seems pretty cool. Figures they'd come out with something like this after I sell my P4P800 rig.
What? You dare question AMD? Fool. You cannot possibly comprehend what powers you toy with.Quote:
Originally Posted by Franky 4 Finger
RELEASE THE PIG.
;) ;)
But personally, I don't think Winchesters live up to their hype... I mean, they're better than NewCastles (although there are many NC's that can break 2.8GHz) but ClawHammer still reigns supreme.
Either way, a P6-architecture rig should blow away most AMD chips in gaming.
How about the Banias? I see them go for $50 on ebay all the time for a 1.4ghz, and with this adaptor for $50 or under that'd be $100 for the combo. Can the Banias overclock at all, or is it not even worth the time and I should just look at the Dothan from the start?
I don't want the P-M to have similar specs to our desktop processors. The processor is trying to be completely different and run at very low FSB and still be very competitive. They should focus on as much performance for as little mhz as possible...........and leave it upto us to jack up the mhz. ;)
The Banias tends to max out around 2ghz. It's just a limitation of the .13 micron design.
theinq talks about p-m EE and 64bit p-m: http://www.theinquirer.net/?article=21666
as far as i understand - pm EE is yonah without low TDP limit.
i really hope that it's compatible with this adapter :)
it wont be...
this combo is EXACTLY what most ppl want for SFF :)Quote:
Originally Posted by Foxie3a
And for everyone who wants to mount something else than a P4 bracket will have some trouble, two of the holes are covered...
:mad:
http://img.hexus.net/v2/internationa...daptor_big.jpg
damn :( then its no point in buying a CT-479..
huh? They're there I think. Just hidden in the photo.Quote:
Originally Posted by Mats
:stick:Quote:
Originally Posted by Xassius
http://www.matbe.com/images/biblio/c...0000004236.jpg
won't you be able to put on most normal heatsinks with the p4 mounting bracket?
Yes you can, but not in this case (some guys here want to use some hardcore cooling that won't use the bracket). As for air cooling, you still have problems. The adapter is about 8 mm high, which makes things very complicated. Almost none of the standard P4 coolers will fit without modding. Would be sweet if they included a higher bracket, but I doubt it since they include a fan that fits. Another problem is the power connector, it's quite much in the way.Quote:
Originally Posted by nathan400
The best thing I can come up with is to use a Zalman 7x00 with the metal pieces upside down, which actually makes it 18 mm higher which is too much. To compensate this you put a piece of metal of right thickness between the sink and the clip, just like some do with other mobiles.
Isn't that bracket in the pic extra high?
^maybe an XP-90 with the retention clips bent up a bit would work?
Watercooling should be ok though right? Adjustable spring tension on most blocks...
Cheers,
Suman
Thats what I was thinking.Quote:
Originally Posted by Revv23
You'd have to be real careful about not crushing the core though...
No overclocking + cooling complications though... might be better just to sell your ASUS board and get an AOpen or DFI socket 479 board... like a lot better...
Looks like they will make it work with Asus P4GD1 (Intel 915P, PCIE, DDR2)
heres some pics I found from THG CEBIT coverage.
http://www.xtremesystems.org/forums/...id=26783&stc=1
http://www.xtremesystems.org/forums/...id=26784&stc=1
http://www.xtremesystems.org/forums/...id=26785&stc=1
More info.
http://www.legitreviews.com/article.php?aid=178
The CT-479 seen above actually has a quite a bit more to it than one might expect. Since the adapter will not allow a standard socket 478 heat sink to be used ASUS provides one with the kit. The blue jumpers are to adjust the FSB (front-side-bus) for CPU support. Depending on which Pentium M is purchased, users can select bus frequency of 533MHz or 400MHz via the jumpers. Also of special note is the power header for the heat sink fan that draws off the socket. Can't say that I have seen that before, but with Intel 755 CPU's only needing 1.34V and the i865/i875 boards offering much higher voltages they should have no problem with the power draw.
To use the CT-479, consumers must first flash their motherboard's BIOS to ensure compatibility. As ASUS refines their motherboard BIOS' to support of this adapter, more ASUS (and maybe other brands) socket 478 motherboards will support this kit. Right now the ASUS P4P800 SE and P4P800-VM are the only motherboards that are supported. The ASUS PC4800E will be the next motherboard to work with the CT-479, and eventually most, if not all, ASUS i865/i875 boards should be supported.
so the voltage controller is onboard so to speak, that sould provide easy vcore moddin' :) however the fact that the FSB is decided by a jumper, not by the mobo makes me wonder if it will be controllable, is the FSB clockgen also on the expansion board ? Perhaps ths too could lead to a mod???? I recall seeing vry high res photos of the board may have a search for them and see if any of the IC's match up...
http://www.x86-secret.com/pics/news/q105/ct479-2b.jpg
Anyone make out the model number on the chip in the top right hand corner, the top left is an op-amp and the two on the right are for GTL/TTL processor signalling. I suspect the one at the top may have somthing to go with the clock
Well you've got more combinations with the FSB jumpers than two. If they just wanted 400 or 533 FSB to choose between they only needed a 1x3 pin jumper. A CPU with the BSel pin removed detects as a 800 FSB CPU by the motherboard, so what happens if you remove one or both of the jumpers? Maybe you won't get any signal to the BSel = 800 MHz FSB... (pure speculation, of course!!)
that would be cool if the chip could run that, odds are it will take a bit more then pulling the jumper to get that thing to boot at 200... we shall see though.
I don't have much experience of overclocking a P4, but what is it like to run a 533 MHz FSB CPU at 800 MHz on a P4P800 with out any mods? Because that's the situation you should get with this adapter if set to 533 and trying to overclock to 800. From the motherboard point of view there are no problems, right? It should be limited by the CPU or adapter AFAIK(either too high FSB, too low Vcore or purely adapter related). Well it could be BIOS related too, we don't know what the new BIOS does when using this CPU.
i just wanna get a low voltage celeron-m 1.4 and change the fsb from 100 to 200, raise vcore to 1.8v, still have a relatively cool setup. whatever board will do this, wether it be a s478 board with an adapter or a s479 board, ddr1 support necessary, i will consider getting. i'm in no rush right now as I am running a nice axp setup :D
Quote:
Originally Posted by Johnny Bravo
It seems to be Chinees or something, there are some weird things written on that chip :confused:
on an IC7-G i was able to get my 1.8a (100mhzFSB) up to 155 on air with an unmodded mobo..Quote:
Originally Posted by Mats
Already posted. I'm not sure that they know what they're talking about. The power connector is not for a fan, it looks like a power input when you look at the install guide (hard to tell, no good pic), which should mean that you probably must make a mod to get higher Vcore. It would have been so much easier if it used mobo power, controlled by BIOS.Quote:
Originally Posted by esoteradactyl
its a floppy plug isnt it
just up 12v rail if so :D
floppy rails are 12volt, ground, ground, 5volt. i'm guessing they are powering the cpu off of the 5volt rail due to the low power required and less of a step down necessary, IF that is indeed a floppy connector. however, in the review linked to above, it says its for a cpu fan. i fail to see the logic of adding a FAN onto the carefully monitored extremel sensitive cpu power lines, i think it would just muck up power distrubtion and possibly cause stability issues. i thnk it looks like a floppy connector but maybe its like the connectors on the old p2/3 slot cpus
I doubt its for the fan, unless its a temperature monitoring output that uses 4-pin like a lot of Hewlitt Packard P2's used to do for the 80mm fan on the back of the case... all that 4-pin did was control the fan RPM depending on the cpu temperature, the actual power was supplied by a 3-pin off the PSU or motherboard.
I've even got a couple of them here... somewhere..
The 4-pin on the adaptor is for a FAN and the extra PIN is for RPM control by the technology of PWM. The new box coolers for Socket 775 has a 4-pin plug as well as the Socket 775 motherboards.
I can't imagine that the FSB is controlled by the adaptor - I think it's simply a identification mechanism (BSEL?) to ensure the CPU won't run overclocked and "safely" boot. How would you lock or change the FSB by shortening something in a CPU socket?... that would need help from the bios wich I don't think is in Asus interest.
I'm not in doubt of the Dothan's WILL be overclockable in the Socket 478 motherboards :)
It's a old press release, but haven't been postet yet: http://www.asus.com/news_show.aspx?id=223 - A official press release from Asus
How do you know that? Yes it got four pins but that doesn't mean it's a fan connector. The new ones used on S775 mobos looks like a 3-pin connector with one extra pin >> the plastic tab is not in the middle (unlike the adapter). I guess they've done this for backwards compatibility.Quote:
Originally Posted by Salkcin
But yeah, it might be a non standard fan connector although I doubt it. Otherwise it's power input with no voltage regulation.... :(
..or wait! If it uses the 5 V line, just switch pins with the 12 V and use a Fanmate to control Vcore.... :stick: Don't think so....
Look at it - the third pin from the upside down has a hole in the platic. If you remember a 3 pin connector on a fan it has a little "tab" in one of the sides - I believe the hole is made for this tab. The 4 pin connector sitting on a FAN is just like a 3 pin, but with extended space for the plastic wall on the motherboard connector so the connector on the adaptor would be "native" 4 pin where a lot of the Socket 775 motherboards has a "native" 3 pin connecter with 1 extra pin.
http://www.legitreviews.com/images/reviews/178/2.jpg
it looks like a floppy plug to me
it is a floppy plug :D
i held this adapter in my owm hands at cebit :)
the jumper is for setting the fsb btw but i think you already knew that :)
so can you further "fine tune ;)" the fsb/cpu clock in the bios still, or will the jumpers limit it? Not much use if you can't clock it!
Like the old Iwill slocket adapter, you're extremely likely going to be limited by the jumpers. So suckit.Quote:
Originally Posted by XeroHouR
well thats quite unfortunate, looks like I'm going with venice lol
no. asus will recognise the cpu as 533 mhz fsb because the board supports celeron (400 mhz fsb) p4 533, and p4 800 mhz. once recognised as p4 533 you will be able (hopefuly) to increase the fsb externaly on 1 mhz increments. and to lock your agp, pci to 33mhz/66mhz. then depends to the ram (good or bad) you can adjust the so called bootstrap - the ratio fsb-ram frequenzy 1:1, 3:4, 4:5 .... and what the current bios can allow you. now about the voltage see this wire mod in the soket: http://forums.legitreviews.com/viewtopic.php?t=1525Quote:
Originally Posted by Mats
hes giving his CPU 1.6v? that gonna kill it i think :eek:
first _real_ test (sort of) of the adapter here: http://forums.legitreviews.com/viewtopic.php?t=1570
so, at least we know now:
the strange plug is indeed a hsf power plug :)
mult and cpu vcore aren't adjustable in bios with adapter :/
They are pulling an all-nighter and i quote:
:slobber: :slobber: :slobber: :slobber: :slobber: :slobber:Quote:
got to look at an advance copy of some of the test data... how can i say im excited with such weak words Very Happy some of the results are very close to ground breaking.
then we'll volt mod it, and do it ourself ;)
oh yeah, someone tell him to do a bandwidth test with that CPU on that board, please
He should be able to do the pin mod too to get more volts to that cpu shouldn't he? Cant be arsed to sign up to forums to post though as its late over here in UK. :)
Can i just confirm that a 1.7 dothan can use dual chanel using this converter or is it a feature only on the 2.13+ dothans.
Thanks
Dual channel is not a feature of the CPU (not for Intel), but a feature of the chipset. So if the chipset and motherboard supports dual channel, you can run dual channel on all Dothans. :)Quote:
Originally Posted by qui
But the 2,13GHz will have better use for it bacause of the higher bandwith between the chipset and CPU. 4,3GB/s vs. 1,7GHz Dothan's 3,2GB/s, but this is what we have overclocking for... ;)
all CPU's hooked on this thing should have it.
ahhh oki I get ya, been an AMD man for some while so was confusing onboard mem controller with what intels northbridge does :confused:Quote:
Originally Posted by Kamerat
is there any word on release date, as i am tearing my amd rig apart this week so would hope to get dothan system running before may.
ok... I was just reading the posts in the thread at legitreviews about the adptor. They guy doing the testing is running at 165mhz FSB / 2.64GHz... so the FSB is adjustable through BIOS, but vcore and multi are not... so basically its like OCing any normal intel system only without vcore adjustment...
Hmm... super pi to 1m places in 28secs though... shiiiiitttt thats amazing! Forgive me if I am mistaken, but I remember seeing (earlier tonight) a 2.6GHz Winchester A64 will only manage around 33s right? These things are FAST... I sure hope AOpen's new i915 board is out when i get back from spring break! cant wait to see what you can do with a Dothan in that board...
yes that sure is fast :)Quote:
Originally Posted by executor252
you need over 4ghz of prescott to get that score :)
hey what original clock is that chip ? 1.6 ?
Thanks
Dan
Results look good though it is disappointing that Vcore cannot be set in BIOS and neither can the multi.
Still a 165 bus and CPU @2600-2700 seems to be close to the limit on stock cooling. I think it is a slight fizz for the product. I was hoping for 2600-2700 BUT with Bus speeds of 200-240.
We will wait and see I guess.
we'll adapt the cooling :DQuote:
Originally Posted by Mikesta
:confused:Quote:
Originally Posted by caater
Really?
-In the pic at Legit reviews forums the HSF is not connected but a cable is connected to the adapter.
-The HSF got one ground, the the cable from the adapter two.
-A fan connector with two gorund pins???
-Look at this pic, now I know it's not a good one. It looks like you connect the adapter to an extender (7 & 8), just like the ones you get together with graphic cards (it's a floppy connector instead though). They connect the HSF wire to the mobo in (10).
Maybe it's possible to use mobo power instead, let's say that it got connection up to the adapter but ends there somewhere. Maybe some measuring and soldering will do the trick. I don't think it's such a good idea to increase Vinput to that adapter, it doesn't look to have any headroom for more juice....
Edit: If you still don't believe me.... it's not a fan plug!
http://www.watch.impress.co.jp/akiba...mage/nctd1.jpg
CT479 ON P4P800-SE(Pentium-M 730 1.6G 533FSB)
http://www.watch.impress.co.jp/akiba...mage/nctd3.jpg
FAN
http://www.watch.impress.co.jp/akiba...mage/nctd4.jpg
SHOW
http://www.watch.impress.co.jp/akiba...mage/nctd2.jpg
OC
http://www.watch.impress.co.jp/akiba...0503/geta2.jpg
http://www.watch.impress.co.jp/akiba...0503/geta3.jpg
http://www.watch.impress.co.jp/akiba...0503/geta4.jpg
http://www.watch.impress.co.jp/akiba...0503/geta5.jpg
AKIBA
Quote:
Originally Posted by Mats
Saaya already said its a fan plug :)
:stick:Quote:
Originally Posted by enzoR
Quote:
Originally Posted by saaya
lmao! i meant floppy. floppy/fan bleh dyslecsic
so what would stop you from taking those stock metal hooks off and putting them on an aftermarket heatsink?
*edit*oh, the power connector, LOL
INTEL DT Pentium-M
NewCode:Expreser
pipeline:13 steps
L1:128K(96+32)
L2:2MB(on die)
FSB:200FSB
Well the results looked promising at the start, but based on the review theres a hue decrease in performance when gaming, although everything else is fairly higher performance wise, but if this adapter decreases performance in gaming which it will mostly be used for its not worth it.
http://www.legitreviews.com/article.php?aid=181
Sounds really nice. :banana:Quote:
Originally Posted by sxs112
Just a dream or reality ?
cu, BitpowerPM
Quote:
Originally Posted by sxs112
OMG :eek: :slobber: :slobber: :wierd:
what socket?
price?
I guess the power cable is for the adapter only, the CPU get its power from the mobo, sweet! Anything else would have been stupid. Someone please try the simple 1.6 Vcore wire socket mod. :D
How will that fix its poor performance in games? unless its cause its running async, but i thought that wasnt a problem with the dothans, perhaps its to do with the way how the 865 and 875 boards works best in 1:1..
I didn't say it would. On the other hand yeah, with more Vcore you'd be getting closer to 1:1 200 FSB, dunno if you can get past it tho. Either Vcore, adapter or CPU limited.Quote:
Originally Posted by qui
What if you can reach higher FSB if the CPU is recognized as 200 MHz like I said earlier?
I read $67USQuote:
Originally Posted by Malachia
Quote:
Originally Posted by Mr. Tinker
near 100€ in italy, i can bet on it! :mad:
Nice pics :up:
So let's get this straight...
-165bus
-No VCORE adjustments
-No Multi adjustments
-Proprietary cooling
-Murders gaming performance
Can you say "SUCK?" :p: The only real use for this is if you are concerned about power usage or were looking for a fast processor to put in a SFF case to run desktop apps.
The new Sonoma Platform should be better for us all. It should at least drop 2/5 of those problems and you could eliminate 2 more with modding. Take another look at the AOpen, it was supposed to be due at the end of this month, but should still be due out soon: http://xtremesystems.org/forums/show...3&page=1&pp=25
Hopefully DFI gets their offering out soon...
It beats the P4 3.73 GHZ EE in Doom3 :eek:
yeah but at 2,5mhz a dothan is almost at the max clock at air....
Dothan does. Dothan on the adapter would be lucky to beat a 1.8GHz chip in D3.Quote:
Originally Posted by Sheik
Hehe. FX55 all the way :D
Quote:
With the CT-479 ASUS has effectively demolished all other desktop Pentium M solutions; there's no reason to even consider a 855GME motherboard from AOpen or DFI, the ASUS solution is cheaper, better performing and is even a much more stable overclocker. Kudos to ASUS for a job extremely well done with the CT-479, it's the only option we'd recommend for those interested in a desktop Pentium M system.
That being said, despite being paired up with enough memory bandwidth, the Pentium M continues to fall behind in desktop performance. As a gaming platform and as a general purpose/office machine the Pentium M does fairly well, but it is in content creation, workstation and media encoding applications that the Pentium M continues to fall behind in. Part of the problem is that the Pentium M needs clock speed to compete, which we saw when we overclocked it up to 2.56GHz. But even at 2.56GHz the Pentium M wasn't a competitive CPU when it came to tasks like media encoding, indicating that if the Pentium M is to succeed on the desktop it's going to need some architectural improvements.
Edit: Okay there is no problems with gaming performance with the adapter in the Anandtech review: http://www.anandtech.com/cpuchipsets...spx?i=2382&p=7
Can you pinmod a voltage increase using this thing?
I wouldn't choose PM for low power because of the price. A64 is a better choice (and 67 W for a 3500+ is still just plain BS).Quote:
Originally Posted by WiCKeD
PM 7x5 : 21 W @ 2100 MHz max
PM 7x0 : 27 W @ 2267 MHz max
T64 ML : 35 W @ 2400 MHz max
T64 MT : 25 W @ 2200 MHz max
mobile A64D : 35 W @ 2000 MHz max (1.35 V)
(Ok you never know if those numbers are made for CPU's that never will show up, just like that 67 W number seems to be.)
That A64D number can easy be translated to desktop CPU's (1.4 V):
3500+: 41.4 W
3200+: 37.6 W
3000+: 33.8 W
...and then there's some headroom for undervolting if you feel like doing that. Much cheaper, more availability, more mobos, and A64E seems to be even more undervoltable = THE SFF choice.
I'm surpised that the new AOpen mobo haven't showed up, one store here in Sweden says (yeah..) they will have it on April 11, thought it would show up earlier in the US.
well, someone please slap that adapter on IC7 now, please :)
Anandtech just reviewed this adapter on an Asus board:
At the end of the article he concluded the P-M was good for everyday desktop tasks and gaming, but it lacked in multimedia and encoding performance. He managed to OC the 2.13 dothan to 2.56ghz (multis become locked when Dothan is used on desktop, and vcore as well). Stock multi of 16 and increased the FSB from 133 to 160.
Just to give you guys a hint at performance, it beats a FX-55 in Doom 3 with High Quality settings. WOW
Click here for complete review.
no bios settings, but it can be changed in windows.Quote:
Originally Posted by Tenknics
so no probs as the clockgen is also available.
...only at 1024x768, with 0.3 FPS, and the FX-55 at stock speed... :stick:Quote:
Originally Posted by Tenknics
that is interesting how it beats the FX-55 in some tests but can't even beat regular a64s in others. Intel has some interesting artitecture here, should be tweaked a bit though with some more content creation, workstation and media encoding instructions because that is where it's seriously lacking.
TBH I question the anandtech numbers as they contradict and differ quite alot from previous numbers posted about the dfi board, their 2.56 dothan is paling ni comparison to other dothans at same speed, plus I was under impression that if you set graphcs to high that its the gpu that takes the brunt not the cpu so that skews numbers even more.
Edit: turns out the reason anandtechs numbers were different is 'cause they have a newer bios that now adds vcore and multis :rehab:
Of course there is a new Beta BIOS from Asus USA for the P4P800-SE offering V-Core up to 1,95 Volt an best gaming performance.
Also the magic FSB200 limit has broken:
http://www.legitreviews.com/images/r.../dothanfsb.jpg
yeah babiiiiie !!!!
Hey guys this is Nate over at Legit Reviews. I just wanted to let everyone here at Xtreme Systems that I have updated my article since I first posted it.
I first tested the board with the ASUS 1008 BIOS revision, but late today ASUS sent me Beta 1009 and it fixes the graphics performance issues, adds voltage support, and the available memory dividers now work. (Before the 1:1 divider didn't run 1:1).
I have spent some more time this afternoon re-overclocking the system and the results thus far are great. I'm running 1.6V on the core with no volt mods with the available options in the new BIOS.
Here is a screen shot of some of the recent testing I have been doing...
http://valid.x86-secret.com/show_oc?id=8208
http://www.legitreviews.com/images/r.../dothanfsb.jpg
I have some of the new BIOS images and more details on my overclocking on this thread.
Thanks for the update Nate! I've updated our front page story with your posting :up:
220FSB on a Dothan, can't wait to see what these chips can do on the right platform! :hehe:
You have to realise that Dothan has 2MB of cache. Its cache also has the lowest latency out there (yup, even lower than A64). Thats why its able to perform very well in apps that benefits from large cache (low latency gives it additional boost). In apps that don't benefit from large cache, its low latency also helps in boosting performance.Quote:
Originally Posted by bh2k
However, Dothan does not have SSE2 nor is the architecture optimised for video apps. Dothan's FPU isn't exactly powerful either (P4 has SSE2 and sheer clockspeed while A64's has the most powerful FPU of any x86 CPU). Have to take note that Dothan is afterall a mobile CPU. If Intel is going to cramp everything P4 has into Dothan, its power consumption will not be at where it is now.
Ok it looks like price wise (since were all going to overclock and put it on water or phase anyway), the dothan adaptor kicks some serious ass. They are going to release a bios for the p4c800-e as well which is great news. So a volt modded board with the vdroop and vmods should allow any voltage issues to be handled, put it under some ln2 or cascade and lets see what they can really do! One question i do have though is which chip would be the one to go for. I.e. a 770 533fsb / 765 400fsb and why? Would a lower spec chip give more???
how do you plan on doing that, the adaptor blocks 2 of the 4 mounting holes :(
great job nate =)...
lol :) what forum is this? do you seriously think for minute that blocking two of the holes will stop the ingenuity of the people on this forum? If so try the pc-world forum :stick:Quote:
Originally Posted by XeroHouR
Getting a 400fsb chip may actually be better for overclocking due to the higher multiplier. Damn man do you have 4 VapoLS's?Quote:
Originally Posted by oublie
the problem is that all 400 fsb chips are all of older stepping and not capable as high fsb as 533mhz ones.. :(
guys with adapter what multis are unloked????
All multipliers from the default one downto 6x are usable. :)
Quote:
Originally Posted by aoc007
No 2 ls' , waterchiller and a 3 stage cascade soon, once i sort some things out :rehab: